https://github.com/gravins/numgraph

Synthetic graph generator
https://github.com/gravins/numgraph

graphs numpy python

Last synced: about 2 months ago
JSON representation

Synthetic graph generator

• Host: GitHub
• URL: https://github.com/gravins/numgraph
• Owner: gravins
• License: mit
• Created: 2021-11-24T09:40:44.000Z (about 3 years ago)
• Default Branch: main
• Last Pushed: 2023-11-07T16:29:11.000Z (about 1 year ago)
• Last Synced: 2024-11-08T00:18:18.599Z (about 2 months ago)
• Topics: graphs, numpy, python
• Language: Python
• Homepage: https://numgraph.readthedocs.io
• Size: 582 KB
• Stars: 12
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
[pypi-image]: https://github.com/gravins/NumGraph/blob/main/docs/source/_static/img/NumGraph_logo.svg

[pypi-url]: https://pypi.org/project/numgraph/



  



# NumGraph

#### Read the [Documentation](https://numgraph.readthedocs.io/en/latest/index.html)

**Num(py)Graph** is a library for synthetic graph generation. The main principle of NumGraph is to be a lightweight library (i.e., ``numpy`` is the only dependency) that generates graphs from a broad range of distributions. Indeed, It implements several graph distributions in both the static and temporal domain. 

## Implemented distributions

### Static Graphs

- Star graph

- Clique

- Two-dimensional rectangular grid lattice graph

- Random Tree

- Erdos Renyi

- Barabasi Albert

- Stochastic Block Model

### Temporal Graphs

- Susceptible-Infected Dissemination Process Simulation

- Heat diffusion over a graph (closed form solution)

- Generic Euler's method approximation of a diffusion process over a graph

## Installation

``` python3 -m pip install numgraph ```

## Usage

```python

>>> from numgraph import star_coo, star_full

>>> coo_matrix, coo_weights = star_coo(num_nodes=5, weighted=True)

>>> print(coo_matrix)

array([[0, 1],

       [0, 2],

       [0, 3],

       [0, 4],

       [1, 0],

       [2, 0],

       [3, 0],

       [4, 0]]

>>> print(coo_weights)

array([[0.89292422],

       [0.3743427 ],

       [0.32810002],

       [0.97663266],

       [0.74940571],

       [0.89292422],

       [0.3743427 ],

       [0.32810002],

       [0.97663266],

       [0.74940571]])

>>> adj_matrix = star_full(num_nodes=5, weighted=True)

>>> print(adj_matrix)

array([[0.        , 0.72912008, 0.33964166, 0.30968042, 0.08774328],

       [0.72912008, 0.        , 0.        , 0.        , 0.        ],

       [0.33964166, 0.        , 0.        , 0.        , 0.        ],

       [0.30968042, 0.        , 0.        , 0.        , 0.        ],

       [0.08774328, 0.        , 0.        , 0.        , 0.        ]])

```

Other examples can be found in ``` test/plot_static.py ``` and ``` test/plot_temporal.py ```.